Inaugural AustMS Math4ML Seminar: Prof Ding-Xuan Zhou

The AustMS Special Interest Group in Mathematics for Machine Learning (Math4ML) is delighted to announce its inaugural seminar. We are excited to welcome Professor Ding-Xuan Zhou from the School of Mathematics and Statistics at the University of Sydney, who will speak on the mathematical foundations of structured deep neural networks.

This seminar launches Math4ML’s online seminar series, which aims to bring together researchers and students interested in the rich connections between mathematics and machine learning. Members of the AustMS community and all interested participants are warmly invited to join us.

Deep learning has been widely applied and brought breakthroughs in speech recognition, computer vision, natural language processing, and many other domains. The involved deep neural network architectures and computational issues have been well studied in machine learning. But not much is known about the modelling, approximation or generalization abilities of deep learning models with network architectures and structures, due to some differences between the classical fully-connected neural networks and structured ones used in deep learning.

An important family of structured deep neural networks are deep convolutional neural networks (CNNs) induced by convolutions. It is believed that local shift-invariance of natural images and speeches plays an essential role in the efficiency of CNNs. Another family of structured deep neural networks are transformers based on attention structures where variable dependence is believed to be crucial in processing natural language data. This talk describes approximation and generalization analysis of deep CNNs, transformers, and related structured deep neural networks.

Mathematics for Machine Learning (Math4ML) is a Special Interest Group of the Australian Mathematical Society. It provides a forum for researchers whose work connects mathematics with machine learning, data science, artificial intelligence, and related areas.